Department of Nutrition and Dietetics, School of Health Sciences, University of Management and Technology, Health Sciences Campus, Raiwind Road, Lahore, Pakistan.
Department of Nutrition and Dietetics, School of Health Sciences, University of Management and Technology, Health Sciences Campus, Raiwind Road, Lahore, Pakistan.
Comput Methods Programs Biomed. 2023 Oct;240:107682. doi: 10.1016/j.cmpb.2023.107682. Epub 2023 Jun 28.
The flaws in dietary assessment methods can generate misleading information and thus may impact on the interventions planned based on that information. Context specific digitalization of dietary assessment tools is a potential way forward to reduce biases and resources involved in data handling.
Two versions of Twenty-Four Hour Recall (24HR) (traditional [24HR Ver-01] and digital [24HR Ver-02]) were tested for data agreement and feasibility by gathering cross sectional paired data on both the versions from 102 participants (18-25 years age). The web based 24HR was setup using the system of Intake24 (New Castle University) with incorporation of South Asian food data base for beverages.
The data sets obtained from 24HR Ver-01 and 24HR Ver-02 on beverage consumption (food items as well as portion sizes) were compared for agreement. The highest percentage of agreement of food item reporting between 24HR Ver-01 and 24HR Ver-01 was during the lunch time. The average kappa value (κ =0.375833) for all the meals indicated a fair agreement betweenVer-01 and 24HR Ver-02 The correlation of portion sizes reported using 24HR Ver-01 and 24 HR Ver-02 was statisticallysignificant for morning snack, lunch and dinner (r = 0.465; r = 0.324; r = 0.407 respectively). According to Bland Altman plot, least agreement between the two versions was found in the portion sizes reported for morning snacks. Data collectors found 24 HR Ver-02 easier in terms of data processing but it was regarded time taking and less convenient by the participants.
The Intake 24 (digital version of 24HR) can be a preferred tool of data collection as the data collected through it may reach fairly good levels of accuracy. Future directions for research like conducting a follow up study with cross over design, expanding the study using food items other than beverages, and testing the digital dietary assessment tool against an objective gold standard of dietary intake can be helpful in reaching more conclusive evidence.
饮食评估方法的缺陷可能会产生误导性信息，从而影响基于这些信息计划的干预措施。针对饮食评估工具的特定情境数字化是减少数据处理偏见和资源投入的潜在途径。
通过对 102 名（18-25 岁）参与者的两种版本 24 小时回顾（传统[24HR Ver-01]和数字[24HR Ver-02]）进行横断面配对数据收集，测试数据一致性和可行性。基于互联网的 24HR 使用纽卡斯尔大学的 Intake24 系统建立，同时包含了南亚饮食数据库中的饮料数据。
对 24HR Ver-01 和 24HR Ver-02 获得的饮料消费数据（食物项目和份量大小）进行比较，以评估一致性。午餐时间食物项目报告的一致性最高。所有餐次平均κ值（κ=0.375833）表明，Ver-01 和 24HR Ver-02 之间存在适度一致性。使用 24HR Ver-01 和 24 HR Ver-02 报告的份量大小的相关性在早餐、午餐和晚餐时具有统计学意义（r=0.465；r=0.324；r=0.407）。根据 Bland Altman 图，两种版本报告的早餐份量大小一致性最低。数据收集员认为，24 HR Ver-02 在数据处理方面更简单，但参与者认为它耗时较长且不太方便。
Intake 24（24HR 的数字版本）可以成为数据收集的首选工具，因为通过它收集的数据可能达到相当高的准确性水平。未来的研究方向可以是进行交叉设计的随访研究，使用除饮料以外的食物项目扩展研究，并将数字饮食评估工具与饮食摄入的客观金标准进行测试，这有助于得出更具结论性的证据。
